In enterprise computing environments where computing resources may be shared between any number of applications, processes, data transactions, etc., the classic problem that frequently arises is how to allocate limited computing resources to various requesting entities which may include, for example, a person, an application, a process, a server and the like. In recent years, this problem has become more manageable with the development of higher power processors and multi-processor computing environments with load-balancing capabilities. However, in high traffic computing environments, maintaining an adequate level of computing resources to adequately manage periods of high volume demand can often be costly. Further, adding additional computing resources is often not cost effective as it may result in extended periods where resources are not sufficiently employed. For example, while all computing resources may be employed during periods of high-traffic which represents 23% of total computing time, many of those assets may be unused for the remaining 77% of total computing time. Efforts, such as load balancing, has resulted in some alleviation of the problems associated with limited resources during high demand, however other problems still exist.
Load balancing is a process by which a load (number of requests, number of users, etc.) is spread throughout a network of servers and other devices in order to prevent individual devices from becoming overwhelmed by traffic, which may cause it to fail or to process the traffic more slowly. Load balancing may also be employed to redirect traffic in the event of a server or device failure thereby providing a level of fault tolerance. While load balancing is effective in managing computing resources in accordance with varying traffic levels, the technology is generally unconcerned with the nature of the traffic itself. In other words, load balancing utilities, as employed within most enterprises, do not apply security controls to computing assets beyond a one-dimensional decision point which is typically data classification.
Therefore, a need exists for a system and method for application of security controls and dynamic provisioning of computing resources that takes into account multi-dimensional decision criteria based on asset classification, business value and resource classification for a requesting entity.